Research On Demand Forecasting For X-Ray Contrast Of B Company In China

Contrast agent is one of the essential drugs in diagnostic imaging and interventional radiology.Although it is a small variety in the statistical classification of drugs in our country,the market share of contrast agent in pharmaceutical industry is growing faster in recent years.As a pioneer entering into Chinese contrast agent market,B company is a leading manufacturer in contrast agent all over the world,whose leading position in Chinese market is affected by generic products made by domestic companies with Chinese market share decreasing continually in recent years.Product demand is hard to be forecasted effectively all the time.Without quantitative forecasting model in sales management,problems of unrealistic demand forecast,a high level of subjective forecast,etc.are easily brought by only using qualitative methods.The forecast accuracy fluctuates and the forecast is biased,which brings about a great challenge to corporate management and product supply department.Through processing of historical sales value of X ray contrast agent and historical sales volume of key SKU A of X ray contrast agent in B company,time serials forecast models as well as multiple linear regression model are developed in this thesis by using relevant forecast theories.Related factors which significantly impact the sales are identified via correlation analysis.To find out the quantitative relationship among them,a multiple linear regression model is set up and used for forecast.Within time serials forecast models set up,there are moving average forecast model,exponential smoothing forecast model and time series decomposition forecast model.Then compared evaluating indexes for each forecast model,suitable models are found out and composed to a forecast models system.Moreover,for tracking and monitoring the forecast quality,signal tracking models are developed for these selected time serials forecast models to ensure the forecast is tracked within the upper and lower limits.Once tracking signal of current using forecast model shows beyond the limits,switching to another forecast model remaining in the limits is necessary for further forecasting.Finally,a models system is set up for this SKU A to improve the forecast accuracy.On the premise of meeting the market demand,it helps reduce and optimize current inventory,guide manufacturing and provide a basis for decision-making for the enterprise.